Bad gas, and its adverse affects on data

I should start off with saying this is a confirmed fix, but
allow me to give you our history of this vehicle.

Type: 2006 Ford Fivehundred 3.0 gas, not E85.

Grandma comes in on the hook that saying her alternator belt
was thrown, car died while driving. Upon inspection, I find
that her A/C compressor locked up- and when drained I found
metal particles inside the oil, which consequently launched
throughout the system. I replaced the belt, a/c compressor,
condenser, dryer, expansion valve, PSI sensor (read 500 psi)
and performed an evaporator core flush.

After everything was said and done I took the car on a test
drive, enough for three ignition key cycles and had a p0171
hard code fault set, along with a pending p0172. It even
started stalling. Hmmm..... well, no one never test drove
the car prior, I jumped and drove the car right into my bay.
Its system KAM was reset from dying on the road after its
belt was thrown off the alternator, so no codes were set.

The fuel trims in the short term were elevated but not
terrible, fluctuating between +8-12% addition, while the
long terms were maxed out at 25 each. There was no
improvement in acceleration, so a vacuum leak was not my
concern. I still had the thought maybe a vacuum line was
knocked off during my a/c repairs so I smoked the intake
system, no leaks. I smoked the brake booster and lines, no
leaks. I pulled the dipstick tube to smoke the crankcase, no
leaks.

On a test drive when the vehicle felt more stable, but still
had lean codes I checked my upstream and downstream o2
values. Keep in mind these are NOT wideband A/F sensors,
these are 0-1v normal operation o2's reporting unburned
oxygen after combustion. The upstreams swung rich to lean
across stoich 200mv- 850 mv, seemingly normal. Only thing I
found strange was while winding out 1st gear on WOT they
read rich as expected, but for a shorter time base than what
I had thought was appropriate, still didn't convince me of
any specific issue. Downstream remained around 650-800mv on
each sensor, reporting a RICH condition on both banks.. not
lean. Weird??

From here I went on and performed my VE testing to see how
the MAF was working and how well air mass was moving through
the system. At idle MAF read around 3.22 g/s seems good (1
gram per liter of displacement). The test passed at 85%
efficiency. MAP readings never exceeded atmospheric
pressure.

I was pretty certain a clogged exhaust was not an issue
since MAP readings at idle were normal, I didn't hear
hissing, there were no misfires, no popping, and honestly,
it didn't feel how a restricted exhaust feels.

Now one piece of information I had purposefully neglected to
mention was that twice, I had an intermittent EGR circuit
code. The EGR has been replaced prior to coming to our shop,
all the inputs were correct, and when grounding the DPFE
signal the engine stalled, everything worked and was fine.
hmmmmm why is this sounding like a bad PCM? PCM's are rare,
we need to prove it isn't our issue...

Finally, before calling a bad PCM I need to check actual
fuel pressure, on this vehicle the manufacturer states:

Key on, engine off (ab) .................... 273 kPa (39.7
psi) Engine running (ab) .................... 273 kPa (39.7
psi)

(a) With a manual gauge, the fuel pressure will be lower.
(b) Based on 60°C (140°F) fuel rail temperature,
at fuel rail temperatures above 60°C (140°F), for
example, hot starts, the fuel rail pressure increases based
on temperature up to 448 kPa (65 psi).

- My scanner read the appropriate 39.7 while manually read
25 psi at prime, and 30 at running" Lower than spec, but it
said it would with a manual gauge.

Now I decide to take a fuel sample, the gas seems slightly
more slimy and darker in color than usual. I then make 10
even marks across my sample bottle with maybe and fill 9
marks of the vehicle's fuel, and add a line worth of water.
Alcohol content is 40%, far higher than the 10% spec needed
for this fuel system! Bad gas! okay we've finally found our
issue.

So if you've read this far, you must think what the hell am
I asking?

My concern is the PCM insisted the command for the addition
of fuel, and we've proven there IS a lean condition, so why
would the O2's all report rich? There was nothing wrong with
these sensors, but it was as if they were lying.. or
confused. I know in strange circumstances like bad gas the
PCM can become confused and do weird things (like set the
EGR code perhaps?) but it was correct to add more fuel.

Only thing I can think of is since there was no freeze frame
data for this car if I had the ability to access maybe, in
fact, the o2's were reporting lean at the time the code set
but I did not capture in the PIDS.

If you've come this far, I thank you for taking on the task
of this long read and I conclude with the question:
Has anyone seen this?

Subscribed to see where this conversation goes.
I'm curious to see what the collective knowledge of the community has to say in this case.
Nice find with the bad gas, that one tends to go under the radar until it bites you a couple of times

I've been down this road before. :silly: On a Saturn Aura 3.6L, AFTER I did chains on it. Filled to the brim with E85.

www.scannerdanner.com/forum/diagnostic-t...anol-fuel.html#22671

My trims behaved about the same as yours, more or less. Sometimes the upstream O2's would stay rich at WOT, sometimes lean, sometimes they'd keep switching. I also ran a VE test and decided fuel was all I had left, because the engine was breathing well otherwise. I'd say your diagnostic process was right on.

I agree with Noah, it isn't always as high up the suspect list as it should be. :lol: I'm much more likely to pull a fuel sample nowadays, if the symptoms fit.

Thank you for the responses, it seems like Tyler had a very similar circumstance as mine, but in a different way. The one idea that recently has been brought to my attention is the possibility of the PCM taking charge of fuel control, working harder to maintain a richer fuel mixture, thus allowing the post 02's to reach rich. If this is true, the PCM works harder, sending a longer PWM command of injectors, thus setting the codes: P0171, P0172. (which as I stated before that I might have been able to recognize in a freeze-frame data situation, but for me that was unavailable). For Fords specifically, and perhaps GM, the fuel quality will be higher up in my diagnostic priority list when it comes to lean on both bank MIL codes. If the theory above is possible, that makes our diagnostic process a little trickier to recognize but is a reminder that fuel quality is as important, if not more important than ever these days.

Also in the future, I will work on equipping photos of my diagnostic approach as others so helpfully do.

-Alexander

I can’t answer that question, so I’m following. But curiosity would make me scope the 02 sensor signal and compare to scan data just to verify that 02 data on the scan tool is correct.

And... nice find!

The combustion products of burning alcohol and gasoline (and indeed LPG/CNG) are different. A convention 0.1v to 0.9v "O2" sensor aka tomato sensor can't cope with that.